# What do our promotions teach customers to expect? > 83% of 175 Black Friday products were the same price or cheaper at least once in the six months either side. What promotions teach customers, and what it costs. Source: https://barkingcat.ai/thinking/what-do-promotions-teach-customers/ ## Thinking · November 2026 · Barking Cat series Every promotion teaches customers a price, and the lesson lasts longer than the promotion. A promotion is judged on what it sold. What it taught arrives next year, after its own report has closed, as customers who wait for the next event and a full price fewer of them pay. A business case that counts only this year’s sales has left that cost out, and in the illustration below it comes to 15% of a line’s contribution. ## What customers learn The research on this is older than most ecommerce platforms. Carl Mela, Sunil Gupta and Donald Lehmann studied more than eight years of household purchases of one frequently bought packaged good, and published the results in the Journal of Marketing Research in May 1997. Their findings were consistent with customers becoming more sensitive to price and to promotions over time, as promotions increased and advertising fell. Customers who were not loyal to a brand became more price sensitive, and in the long run, the authors found, price promotions made loyal customers more sensitive to price as well. Customers also learn the calendar. Which? tracked 175 products on offer at eight large retailers through the four-week Black Friday window of 2024, and reported on 25 November 2025 that 83% were the same price or cheaper on at least one occasion in the six months before or after. None was at its lowest price of the year on Black Friday itself. My reading, not Which?’s: a customer who waits for the next event is right to, most of the time. You cannot teach a cat to bark. You can teach it to come running at the sound of the tin opener. ## The working The numbers below are an illustration. Take a line that sells 10,000 units a year at a full price of £100, with a contribution margin of 50%: £50 a unit after the cost of the product and the cost of serving the order. It makes £500,000 a year. Now suppose a run of promotions teaches three in ten of its customers to wait for 25% off. Next year the line sells exactly the same 10,000 units. Seven thousand sell at £100 and make £50 each. Three thousand sell at £75 and make £25 each. The line has lost £75,000 of contribution while selling the same number of units. Measured on volume, it is holding up. The promotions that taught the lesson may each show a profit in their own reports, which closed months earlier. ## When the price was never repeatable A cheap purchase adds a second cost. Distressed stock can be sold cheaply and still make a healthy margin, because it was bought cheaply. Suppose a line’s normal unit cost is £50, a job lot of the same kind of stock cost £25, and it sold at £70. Each unit made £45. The customer learned £70. When the job lot runs out, the business buys at £50 again, the customer still expects £70, and each unit makes £20. To earn at the learned price what it earned on the job lot, the business has to sell two and a quarter times as many units. Under that sits an accounting problem. For as long as the cheap stock lasts, the P&L reports a margin that belongs to one purchase rather than to the way the business trades. I have seen this in practice: a job lot traded for three years and never separated from the rest of the stock, whose margin the board read as the margin of the business until the stock ran out. ## A lesson or a hangover? Weak full-price sales after a promotion do not prove that customers learned anything. Two other explanations come first, and each leaves a different mark. The first is pull-forward: customers bought in the promotion what they would have bought in the following weeks. It shows as a trough that recovers within weeks, and the [Promotion Economics workbook](https://barkingcat.ai/thinking/promotion-economics/) already counts it. The second is the market, or a competitor’s prices, which hit customers who never bought on promotion as hard as the ones who did. A learned price leaves its own mark. It lasts beyond the trough, and it shows most among the customers who bought on promotion: their share of purchases made at full price falls and stays down, while other customers carry on as before. If that share returns to its old level within a season, it was a hangover rather than a lesson. ## The law has noticed as well A full price that is rarely charged is also a legal exposure. On 30 July 2026 the High Court gave judgment in the Competition and Markets Authority’s case against Emma Sleep, an online mattress retailer, which had admitted some breaches on reference pricing. The court rejected the CMA’s proposal that a fixed ratio of sales, one at full price for every two on discount, must stand behind a reference price. It said a lopsided ratio can still be a warning sign. I am not a lawyer, which the legal profession will regard as the least contentious sentence in this piece. The point for a board is narrow: if the business cannot show who paid the full price, it should expect to be asked. ## What to do first Measure it, because measuring is cheap. Two numbers answer most of the question. The first is the share of this year’s sales made at full price, against the same figure two years ago. The second compares customers who bought in this year’s Black Friday with customers who did not, on the share of their later purchases made at full price. Both come from order data the business already holds. Then judge the next sale on what it teaches as well as what it sells. The expensive option is to stop promoting altogether. The cheaper middle route is to hold one category out of the next blanket promotion, alongside a similar category that stays in, and read both. The Promotion Economics workbook will put a number on what that category gives up before you start. ## How you will know The measure is the full-price share of sales in the held-out category against the matched one. The guardrail is total contribution, so that a category which holds its price but loses its customers does not count as a win. Match the two categories for season and stock, and keep the site-wide promotions the same for both. Read them at the end of the sale and again twelve weeks later. Agree the rule before you start: if the held-out category’s full-price share is no better and its contribution is lower, put it back in. ## The decision The decision for the board is whether next year’s promotion plan is approved on what it will sell, or on what it will sell and teach, and who owns the second number. Two questions for Monday. What share of our sales was made at full price this year, and two years ago? Which category could we hold out of the next sale? This piece rests on one retailer’s experience, a packaged-goods study from 1997 and a consumer group’s price tracking. It cannot tell you how quickly your own customers learn a price or how long they take to forget one, and for products bought once every few years I have not found anyone who has measured either. ## Sources - Carl F. Mela, Sunil Gupta and Donald R. Lehmann, [The Long-Term Impact of Promotion and Advertising on Consumer Brand Choice](https://www.slideshare.net/slideshow/salee/15155660), Journal of Marketing Research, vol. 34, May 1997, pages 248 to 261 - Which?, [Most Black Friday deals the same price or cheaper at other times of the year](https://www.which.co.uk/policy-and-insight/article/dont-believe-the-hype-most-black-friday-deals-the-same-price-or-cheaper-at-other-times-of-the-year-which-finds-a3TbD5x5KADs), 25 November 2025 - CMA v Emma Matratzen GmbH and others, High Court, 30 July 2026, as reported by [Lewis Silkin](https://www.lewissilkin.com/insights/2026/08/03/uk-consumer-law-revolution-reference-prices-emma-sleep-case-update-high-cour-102nfr0) and [Legal 500](https://www.legal500.com/intelligence/united-kingdom/consumer-protection/high-courts-emma-sleep-judgment-offers-insights-into-misleading-reference-prices); [CMA case page](https://www.gov.uk/cma-cases/emma-group-consumer-protection-case) ## About the author Steve Webster is an independent adviser to the boards and investors of UK retail and consumer businesses from £20m to £1bn: an operator’s judgement on ecommerce, technology and AI, before the board commits the capital. Formerly Global Ecommerce Director at Barbour, with 27 years running retail P&Ls up to £240m. ## Contact steve@barkingcat.ai Barking Cat Advisory Limited. Registered in England and Wales, company number 17338003.